Naltrexone has reduced craving and relapse in some, but not all, alcoholics. One potential variable that might predict naltrexone, and potentially other opiate antagonist responses, is a difference in a single nucleotide substitution in the mu opiate receptor protein that makes that receptor respond differently to endogenous beta-endorphin as well as to exogenous opiate antagonists like naltrexone. An (A)denine to (G)uanine substitution at position 118 (A118G) in the coding region of the mu opiate receptor accounts for an asparagines (asn) to aspartate (asp) amino acid substitution at position 40 (asn40asp) in the receptor protein that occurs in about 15% of the population. The purpose of this1 proposal is to examine the role of this allelic difference in naltrexone responsivity in a well-established sub-acute dosing, brain imaging, and bar-lab paradigm. Interaction with another functional allele difference (va!158met substitution) in catechol-omethyl- transferase (COMT), an enzyme that controls CNS dopamine tone, will be explored. Three hundred non-treatment seeking alcoholics will be assessed and subtyped for mu opiate receptor and COMT allelic variants. Eighty individuals (40 with the more common AA gene and 40 with either an AG or GG gene) will be randomly assigned to take either naltrexone (50 mg/day) or a matching placebo for 7 days. Val and Met alleles of the COMT gene will be equally distributed by urn randomization to all groups. After 5 days of natural drinking and one day of abstinence, subjects will undergo an alcohol cue-induced fMRI brain scan on day 6 of study drug. On day 7, after a standard alcohol drink, stimulation, sedation, intoxication, and craving will be evaluated over 40 minutes. Then subjects may choose to drink up to 8 minidrinks over a 2-hour period. It is hypothesized that when taking naltrexone, AG/GG (40asp) subjects compared to subjects with the AA (40asn) will show a greater reduction of cue-induced brain stimulation in the nucleus accumbens, less alcohol induced stimulation, and less free choice drinking. Those with both the asn40asp and val158met substitutions could have the strongest responses to naltrexone. This data will provide support for the use of genotyping in predicting which treatment seeking alcoholic will or will not respond to naltrexone. Also, this work brings pharmacogenetics to our Center for future medication/gene interaction studies.